Residency · Residency · Nuclear Medicine

Gastrointestinal Bleeding Scintigraphy

Overview

Tc-99m red blood cell scintigraphy detects active gastrointestinal bleeding at rates as low as 0.1 to 0.4 mL per minute, making it substantially more sensitive than catheter angiography, which requires bleeding rates of 0.5 to 1.0 mL per minute. The primary role of the study is to localize the bleeding site so that angiographic intervention or surgery can be directed appropriately. It is most useful for evaluating occult lower GI bleeding when endoscopy has been non-diagnostic.

Radiopharmaceutical: Tc-99m Labeled Red Blood Cells

Labeling Methods

Three methods exist for labeling red blood cells with Tc-99m, and the choice of method significantly affects study quality. In vitro labeling using the UltraTag kit provides the highest labeling efficiency at greater than 97% and is the preferred method. Blood is drawn, the red blood cells are separated, stannous ion is added, and Tc-99m pertechnetate is incubated with the cells before reinjection. This method produces the least free pertechnetate and therefore the fewest false positives from thyroid, stomach, and kidney activity. Modified in vivo labeling offers intermediate efficiency of approximately 85 to 90%. Stannous pyrophosphate is injected intravenously, blood is then withdrawn, Tc-99m is added to the blood sample in vitro, and the mixture is reinjected. In vivo labeling is the simplest approach but the least reliable, with only 75 to 80% efficiency. Stannous pyrophosphate is given intravenously followed by intravenous Tc-99m pertechnetate. The higher proportion of free pertechnetate causes gastric, thyroid, and renal uptake that can produce false positives, so this method is not recommended for GI bleeding studies.

Dose

The administered dose is 20 to 30 mCi (740 to 1110 MBq) of Tc-99m labeled red blood cells.

<image>Tc-99m RBC GI bleeding scan showing progressive accumulation of tracer in the right lower quadrant at 25 minutes, tracking in an antegrade pattern through the ascending colon consistent with a cecal bleeding source</image>

Technique

Dynamic Imaging

The study begins with continuous dynamic acquisition using 1-minute frames for 60 to 90 minutes, with the camera positioned over the abdomen in the anterior view. Immediate flow images at 2 to 5 second frames during the first minute may detect very active bleeding.

Delayed Imaging

If dynamic imaging is negative, delayed static images are obtained at 2, 4, 6, 12, and up to 24 hours. Because the red blood cells remain labeled in the blood pool, intermittent bleeding that occurs hours after injection can still be detected. However, delayed images are less reliable for precise localization because blood moves through the bowel over time.

SPECT/CT

SPECT/CT can be performed when planar imaging shows equivocal activity. It improves localization by distinguishing bowel activity from vascular structures and is particularly useful in patients with complex anatomy such as post-surgical changes or a redundant colon.

Interpretation

Positive Study

A positive study shows a focal area of increased activity that was not present on earlier images. This activity increases in intensity over time and moves in a pattern consistent with bowel peristalsis, either antegrade or retrograde. It must appear in an area where no blood pool activity is expected, specifically not overlying major vascular structures.

Localization

Accurate localization requires tracking the movement pattern of tracer across sequential frames to identify the site of origin. Antegrade movement indicates that the bleeding source is proximal to the direction of travel. Retrograde movement can occur when small bowel bleeding refluxes or when blood tracks retrogradely through the colon. The earliest frame showing focal activity is the most reliable indicator of the bleeding site.

Cine Display

Images should always be reviewed in cine (movie) mode to identify the first appearance and subsequent movement of the bleed. This is essential for accurate localization, as static images alone can be misleading. The interpreter should look for the "blush," which is the earliest appearance of extravascular activity.

Sensitivity Thresholds

Bleeding rates as low as 0.05 to 0.1 mL per minute can be detected on dynamic imaging, compared to the 0.5 to 1.0 mL per minute threshold required for catheter angiography. Total blood loss as small as 3 to 5 mL may be visible. The sensitivity for detecting active bleeding ranges from 85 to 95%. However, localization accuracy is lower than detection accuracy, at approximately 70 to 85%.

<image>Serial 1-minute frames from a GI bleeding scan displayed in cine format, showing initial focal activity in the splenic flexure region at 18 minutes with subsequent antegrade movement through the descending and sigmoid colon</image>

False Positives and Pitfalls

Causes of False Positive Localization

Free pertechnetate from poor labeling can produce uptake in the stomach, thyroid, and kidneys, mimicking a bleeding source. Using in vitro labeling minimizes this problem. Genitourinary activity from the bladder, ectopic kidney, or ureter can simulate bleeding. Vascular structures including the aorta, iliac vessels, varices, aneurysms, and hemangiomas may be confused with a bleed. Accessory spleens or splenosis can produce focal uptake mimicking bleeding. Penile blood pool activity in male patients and menstrual uterine activity in female patients are additional potential sources of confusion.

Localization Errors

Blood tracking through the bowel can be misidentified as originating from a downstream location rather than the true source. Retrograde movement of blood in the colon is common and can mislead localization. Delayed images obtained beyond 4 to 6 hours are unreliable for precise localization because blood redistributes extensively through the bowel. Small bowel bleeding is particularly difficult to localize accurately.

Clinical Decision Making

Positive Scan

A positive scan guides selective mesenteric angiography for confirmation and potential embolization. If angiography is negative because the bleeding has stopped, the scintigraphic localization still provides valuable information for surgical planning. A positive scan in a patient with ongoing hemodynamic instability warrants urgent angiographic intervention.

Negative Scan (No Bleeding Detected)

A negative dynamic study has strong negative predictive value, with fewer than 1% of patients with negative dynamic imaging requiring urgent surgery. The patient can be observed, and repeat imaging may be considered if clinical bleeding recurs. Alternative causes of GI blood loss, such as an upper GI source or coagulopathy, should be explored.

Timing of Study

The study is most valuable when performed during active bleeding, as evidenced by positive stool guaiac, hemodynamic changes, or a dropping hemoglobin. Initiating the study promptly when bleeding is suspected maximizes the detection rate. Studies started hours after bleeding has ceased have a substantially lower yield.

<image>False positive on GI bleeding scan: free pertechnetate activity in the stomach and kidneys due to poor RBC labeling efficiency with in vivo method, mimicking upper GI bleeding</image>

Comparison with Other Modalities

CT Angiography (CTA)

CT angiography is rapidly available and provides high spatial resolution. It requires active bleeding at the time of the scan, with a sensitivity threshold similar to catheter angiography at approximately 0.3 to 0.5 mL per minute. It cannot detect intermittent bleeding but can provide alternative diagnoses such as tumor, diverticulosis, or vascular malformation. CTA is increasingly used as the first-line study in acute GI bleeding.

Catheter Angiography

Catheter angiography offers therapeutic capability through embolization but requires a higher bleeding rate for detection, approximately 0.5 to 1.0 mL per minute. It is invasive, involves radiation exposure and contrast risk, and is typically used after a positive scintigraphy or CTA study for targeted intervention.

Endoscopy

Endoscopy is the first-line approach for upper GI bleeding and many cases of lower GI bleeding, offering both diagnostic and therapeutic capabilities. Its utility is limited when bleeding is rapid or diffuse, as visualization becomes poor.

Clinical Pearls

Always use in vitro RBC labeling for GI bleeding studies to minimize free pertechnetate and false positives.

Cine review is essential. Never rely solely on selected static images for localization.

The earliest frame showing focal activity is the most reliable indicator of the bleeding source.

Delayed images beyond 4 hours should be used only to confirm ongoing or recurrent bleeding, not for precise localization.

A negative dynamic study has high negative predictive value and supports conservative management.

Small bowel bleeding is the most challenging to localize accurately. A tagged RBC scan combined with CTA may provide complementary information.

The study should be initiated as soon as possible when active bleeding is suspected. Delays reduce sensitivity.

References

• SNMMI Procedure Standard for GI Bleeding Scintigraphy.
• Defined Role of Tc-99m RBC Scintigraphy in GI Bleeding. Seminars in Nuclear Medicine, 2016.
• Dam, H. Q., et al. "GI Bleeding Scintigraphy Practice Recommendations." Journal of Nuclear Medicine, 2014.
• ACR Appropriateness Criteria: Radiologic Management of Lower GI Tract Bleeding.

Read this lecture as Markdown